Actions

Understand what actions are and and how to work with them.

Actions are executable instructions that are delivered to end-users (real or virtual humans), often through an interface (an app, a screen, speakers, and more). They serve the purpose of helping you implement personalization in your solutions.

During the course of running a solution, different actions are delivered at different moments in different ways to the targeted users (users who meet specific criteria).

For example, the solution plays a personalized audio greeting ("Hi Jamie!") whenever the user (Jamie) is detected by a camera through facial recognition.

You can use logics to define the conditions of action delivery for your solution.

Based on the configured logics, Affective Computing (powered by Virtue) identifies all the profiles for whom an action must be delivered and the best moment to deliver them.

How Affective Computing (powered by Virtue) selects actions for delivery

Action selection is powered by the following components of EDAA™, the underlying technology of Affective Computing (powered by Virtue):

  • Content Prediction Model: A model that predicts the most suitable content to be delivered for any given user.

  • Interaction Automation System (IAS): Automation of interactions to ensure seamless and relevant communication with each user.

Action personalization

You can specify whether an action must be delivered as-is (without modification) or generatively improved by EDAA™, the underlying technology of Affective Computing (powered by Virtue).

You can tailor generative actions to your users' psychological profiles and align them with their current mood. (To learn more about how non-literal moods are defined and handled by Affective Computing (powered by Virtue), see Mood identification). Incorporating personalization ensures that your solution delivers engaging and relevant actions for each user.

Action categories

In Affective Computing (powered by Virtue), you can work with the following action categories:

Action category
Description
Action types

Content

Using Content actions, you can deliver media content, such as images, videos, sound, visual effects, and more.

Note: Content actions usually do not require a response from users.

As a solution designer, you can specify the actual content to be delivered. For example, if you want the solution to display an image, you can enter the link to the image when declaring (creating and configuring) the action.

You can declare content actions of the following types:

  • Sound

  • Visual effects

  • Geolocation

  • Acceleration

  • Music

Interactions

Using Interaction actions, you can ask questions or deliver messages as statements. Note: Interaction actions usually require a response from a user.

You can declare interaction actions of the following types:

  • Statements

  • Questions

Triggered actions

Using Triggered actions, you can deliver customized actions in response to changes in a user's external environment. The action is delivered only when the defined values or value range exist for the specified parameters.

For example, adjusting music, sending alerts, or modifying elements in a video game or app in real time based on the user's mood.

Triggered actions do not need to be classified by type.

Within a project, related or similar actions can be grouped and organized using attributes.

Creating and managing actions in projects

You can create and manage actions using the Portal, our no-code SaaS GUI platform, or through a low-code integration with our APIs.

Working with actions using the Portal

Feeding actions and attributes

Working with actions using the APIs

Working with actions and attributes
Action and attribute feeding

Last updated

Was this helpful?